There are two types of research- primary and secondary. The difference between them relates to whether the research is original to the researcher(primary) or whether it came from another source(secondary). Within these caregories, information can either be internal-from inside the organisation, or external- from outside the organisation.

Primary research is data and information that the business has gathered first-hand and has not been gathered before. 

Internal primary research data sources include:

• Sales figures for the business' own products

• Customer data held on a central database

External primary research methods include:

• Questionnaires and surveys such as online feedback forms

• Interviews and focus groups

• Mystery shoppers and other observation techniques

• Trials of prototypes (An initial version or mock up of a concept for further development) or new products for gathering feedback.

It is more cost effective for a business to conduct secondary research before it starts conducting primary research. This allows it to build an understanding of the market and identify any major barriers before committing to expensive research. It also allows the business to develop some assumptions (for example, people interested in Sky may also be tempted by Sky bundles and other packages), which means the business can use the primary research to test its assumptions. This helps to restrict the scope of the primary research, which can be expensive and is rarely able to answer all questions about a particular market.

Secondary research uses data and information that has been collected before either from within the organisation or externally. This is sometimes referred to as desk research and sources include:

• Reports from sales and regional representatives

• Previous market research

• Trad journals and websites

• Books and newspapers

• Industry reports from industry associations and government departments

• Census data and public records

Internal examples include business data on customers and financial records which can be obtained from loyalty cards such as those used by Tesco, Superdrug and Waitrose. Loyalty cards capture sales records which identify customer preferences. These enable a business to market similar products to those a customer frequently purchases, or issue vouchers to regain sales which may have lapsed.

External examples include commercially published reports such as those produced by Thomson Reuters on market data around the world., government statistics which provide information about market trends, UK population and demographic data to inform ethical decisions.

Other examples include trade journals such as Which? magazine providing information on consumer rights and expert advice about products and services, and media sources such as news articles and consumer voices via social networking sites.

However businesses need to be aware of the limitations of secondary research. For example:

• Information may be old

• Information may be biased to promote a particular cause

• Collection methods may have been unreliable.

Research methods can be qualitative, quantitative or a mixture of both. Well-planned market research often involves a combination as they can reveal different things about the same market.

No matter how small or large a market research project may be, any type of research performed poorly will not give relevant results. All research, no matter how well controlled, carries the potential to be wrong which is why stringent tests are necessary to ensure claims are not made without rigorous testing.

There are many reasons why research may not provide trustworthy results and one example is whether the research is really measuring what it claims to be measuring. To ensure the integrity of the data, certain measures must be in place to carry out suitable checks. Each of these following measures needs to be applied in conjunction with one another.

Validity- This is the process of ensuring data collected are valid by using original sources or tracing sources back to their original point. 

Reliability- Validity questions whether the research measured what it intended to measure and must be reliable. Reliability does not necessarily imply validity; a reliable measure need not be valid. Reliability is chiefly concerned with making sure the method of gathering leads to consistent results.

Appropriateness- Is the data collect appropriate for the task at hand? Is the data collected ethical? Are the methods uses to collect the data appropriate?

Currency- This is a term used to describe how up-to-date or current the research is. Can the data be collected and processed in a time efficient manner.

Cost- The cost of data collection must also be considered to ensure it is within the budget of the business. Market research often takes time to devise and conduct research activities and often travel and accommodation costs apply to the person conducting the research.

Quantitative data is comprised of numbers which can be analysed mathematically and or presented as a graph. This data can include sales figures, market values and so on. It is often quicker to collect and interpret this data as computer models can be used to analyse it. This can be used to produce objective (a judgement which is not influenced by personal opinions or points of view, neither biased nor prejudiced, and can be validated) results.

Qualitative data are subjective (based on personal opinions) and often open-ended text responses. It can involve interviews with customers or focus groups or invite respondents to  add reasons or comments to surveys. Results usually provide a wide range of answers based on personal experience, feelings and expectations.

The types of data gathered and the method or tools used can skew research results. Research findings should be based on sufficiency of data collection.

How were customers selected?

Why did other customers not respond?

What would customers say at other times of the year?

How many customers chose not to respond?

What did they like about it?

All research should have a focus which is related to the purpose for conducting the research and what you want from the study. Research is not valid if it arrives at a result but is unable to prove anything. However, research outcomes can occur as a result of chance, such as some of the discoveries by Stephen Hawking, for example that black holes should emit radiation.

You have been exploring some of the methods which can be selected for collecting and gathering data but the results from analysis also undergo a selection process. For example, in the skin product example showing the limitations of the information presented to the public, questions should be asked about the data not being used and how it can also skew results.

 When undertaking any form of research, considerable amounts of data are often collected and a decision needs to be made about which to use and what to leave out and why. The clearer the focus of the research and the comprehensiveness of the planning, the less likelihood there is of becoming overly bogged down in data. This is particularly so when trying to analyse qualitative data.

Extracting meaning from data relies on time and possibly the use of various tools to help with the process. Again it is important to plan for these activities. For larger scale studies researchers might purchase assisted qualitative data analysis software. Although such software helps to identify trends and patterns in responses, the analysis still relies on human interpretation to extract meaning from the data.